Waiting Room Prompt Chinese Speech Recognition Audio Dataset

In modern hospitals, managing waiting room information and enhancing patient experience remains a challenge. Existing speech systems are mostly limited to simple dialogues or fixed responses, making it hard to handle complex oral inputs. The construction of the Waiting Room Prompt Speech Recognition Audio Dataset aims to provide more accurate speech recognition models and offer more efficient waiting room management solutions for hospitals and clinics. Data is collected in real hospital environments using various recording devices, including directional microphones and portable recording equipment, to ensure diversity in background noise. The data undergoes multi-round annotation and expert review, with an annotation team comprising linguistics experts and medical practitioners, exceeding 20 members. The data is subjected to noise filtering, speech enhancement, and other preprocessing steps before being stored in MP3 format, organized by speaker, scenario, and other labels.The core advantage of this dataset lies in its high-quality annotation accuracy, achieving over 95% consistency and completeness. It innovatively employs voice noise filtering and enhancement technology to accurately simulate real hospital environment usage conditions. In terms of application value, it can reduce the workload of medical assistants and improve patient experience; compared to similar datasets, it offers unique support for professional medical terms and dialects, with its scarcity reflecting in the difficult-to-obtain real hospital environment recordings. It is also suitable for other high-noise environments, such as large customer service centers, and provides good scalability and versatility.